Deletions of the long arm of chromosome 20 [del(20q)] are observed frequenTly in malignant myeloid disorders, but rarely in lymphoid neoplasms. To identify a commonly deleted segment, we examined the breakpoints in 74 patients who have myeloid disorders characterized by a del(20q); however, we were unable to define such a region by cytogenetic analysis alone. By using fluorescence in situ hybridization (FISH) of unique sequence probes that map along the length of 20q, we have delineated the smallest commonly deleted segment of this chromosome. This deleted segment extends from 20q1 1.2-12, is flanked by RPN2 proximally and D20S17 distally, and contains the SRC, P107, and ADA genes. These results suggest that recessive mutations involving a tumor suppressor gene that is located within the commonly deleted segment may play a role in the pathogenesis of this disease. We now propose to use cytogenetic and molecular approaches to map the commonly deleted segment, and to identify and isolate the gene(s) within this region that is involved in leukemogenesis. The initial step will be to narrow the commonly deleted segment by using cytogenetic mapping techniques, such as FISH, to examine leukemia cells characterized by a del(20q) or translocations involving 20q11.2-12. leukemia cells will be examined for mutations of candidate genes that map within the commonly deleted segment, such as the P1O7 gene, which encodes an RB1-like protein. Several experimental techniques will be used to search for previously unidentified expressed sequences located within the commonly deleted segment and, if appropriate, to determine whether mutations of these genes are present in myeloid leukemia cells with rearrangements of 20q. If mutations of a gene on 20q are identified, we will (1) determine the spectrum of mutations involving the gene in myeloid leukemia cells with a del(20q), (2) determine whether such mutations are present in myeloid leukemia cells (or lymphoid leukemias) which do not have abnormalities of this chromosome, and (3) identify the consequence(s) of the mutations on the function of the gene/protein. Although a number of tumor suppressor genes have been characterized in the solid tumors, such genes have not yet been identified in the hematologic malignant diseases (a candidate gene, pl6,which may be involved in lymphoid leukemias has recently been identified). We hope that by examining recurring deletions, we can begin to evaluate the role of recessive mutations in the pathogenesis of leukemias.